GLM-4.6 vs Kimi K2 Instruct 0905 vs Relace Apply 3
Too close to call on our weighted score (GLM-4.6 41, Kimi K2 Instruct 0905 38, Relace Apply 3 32). The right pick depends on what you value most.
Z.ai (Zhipu)GLM-4.6
41/100- ECI—
- Price$0.60 / $2.20
- Context205K
Moonshot AIKimi K2 Instruct 0905
38/100- ECI—
- Price$0.60 / $2.50
- Context262K
Relace
Relace Apply 3
32/100- ECI—
- Price$0.85 / $1.25
- Context256K
Too close to call
It is close. Our weighted score puts them within 3 points (GLM-4.6 41/100, Kimi K2 Instruct 0905 38/100, Relace Apply 3 32/100), so choose by what matters most for your work: Relace Apply 3 on price and Kimi K2 Instruct 0905 for long inputs. The score weighs price 50%, inputs & features 30%, context window 20%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.
- CapabilityNot enough dataNo independent benchmark covers every model here yet
- Lowest priceRelace Apply 3Relace Apply 3 $0.95 · GLM-4.6 $1.00 · Kimi K2 Instruct 0905 $1.07 per 1M tokens (3:1 blend)
- Longest contextKimi K2 Instruct 0905Kimi K2 Instruct 0905 262,144 · Relace Apply 3 256,000 · GLM-4.6 204,800 tokens
- Widest inputsSame inputsGLM-4.6: Text · Kimi K2 Instruct 0905: Text · Relace Apply 3: Text
- Self-hostingGLM-4.6 and Kimi K2 Instruct 0905Publishes downloadable weights (Modified MIT)
| Measure | Weight | GLM-4.6 | Kimi K2 Instruct 0905 | Relace Apply 3 |
|---|---|---|---|---|
| Price | 50% | 48 | 46 | 49 |
| Inputs & features | 30% | 35 | 25 | 0 |
| Context window | 20% | 32 | 37 | 36 |
| Overall | 100% | 41/100 | 38/100 | 32/100 |
Left out because at least one model lacks the data: capability. The remaining weights were rescaled.
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | GLM-4.6Z.ai (Zhipu) | Kimi K2 Instruct 0905Moonshot AI | Relace Apply 3Relace |
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | — | — | — |
| ECI rank | — | — | — |
| Price per million tokens | |||
| Input | $0.60 (best) | $0.60 (best) | $0.85 |
| Output | $2.20 | $2.50 | $1.25 (best) |
| Cached input | $0.11 | — | — |
| Blended (3:1) | $1.00 | $1.07 | $0.95 (best) |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Official Z.AI API | Median of 12 providers | Median of 3 providers |
| Limits | |||
| Context window | 204,800 tokens | 262,144 tokens (best) | 256,000 tokens |
| Max output | 131,072 tokens (best) | 98,304 tokens | 128,000 tokens |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | No | No | No |
| PDFs | No | No | No |
| Audio | No | No | No |
| Video | No | No | No |
| Reasoning | Yes | No | No |
| Tool calling | Yes | Yes | No |
| Structured output | No | No | No |
| Availability | |||
| Weights | Open | OpenModified MIT | Proprietary |
| API model ID | glm-4.6 | — | — |
| API providers | 29 (best) | 13 | 3 |
| Released | Sep 30, 2025 | Sep 5, 2025 | Sep 26, 2025 |
| Knowledge cutoff | Apr 2025 | — | — |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
- GLM-4.6$10.40
- Kimi K2 Instruct 0905$11.00
- Relace Apply 3$11.00
Which should you choose?
Which is better: GLM-4.6, Kimi K2 Instruct 0905 or Relace Apply 3?
It is close. Our weighted score puts them within 3 points (GLM-4.6 41/100, Kimi K2 Instruct 0905 38/100, Relace Apply 3 32/100), so choose by what matters most for your work: Relace Apply 3 on price and Kimi K2 Instruct 0905 for long inputs. The score weighs price 50%, inputs & features 30%, context window 20%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.
Which is cheaper, GLM-4.6, Kimi K2 Instruct 0905 or Relace Apply 3?
Relace Apply 3 is cheaper at $0.85 input / $1.25 output per million tokens (median across 3 API providers). GLM-4.6 costs $0.60 input / $2.20 output per million tokens (official Z.AI API price); Kimi K2 Instruct 0905 costs $0.60 input / $2.50 output per million tokens (median across 12 API providers). At a typical mix of three input tokens to one output token, that is $0.95 per million tokens for Relace Apply 3 versus $1.00 for GLM-4.6 (1.1× as much) and $1.07 for Kimi K2 Instruct 0905 (1.1× as much).
Which scores higher on benchmarks?
There is no independent benchmark that covers all three models yet. GLM-4.6 has not been scored yet, Kimi K2 Instruct 0905 has not been scored yet and Relace Apply 3 has not been scored yet.
Which is better for coding?
There are no published SWE-bench Verified results for GLM-4.6, Kimi K2 Instruct 0905 and Relace Apply 3 yet, so there is no like-for-like coding score. Test both on a sample of your own tasks. Note that Relace Apply 3 does not support tool calling, which most coding agents need.
Which has the bigger context window?
Kimi K2 Instruct 0905 has the largest context window at 262,144 tokens, against 256,000 for Relace Apply 3 and 204,800 for GLM-4.6. Maximum output per response: GLM-4.6 up to 131,072, Kimi K2 Instruct 0905 up to 98,304, Relace Apply 3 up to 128,000 tokens.
Which can read images, PDFs, audio or video?
GLM-4.6 accepts text; Kimi K2 Instruct 0905 accepts text; Relace Apply 3 accepts text. They handle the same number of input types.
Are any of these open source?
GLM-4.6 and Kimi K2 Instruct 0905 publishes its weights (Modified MIT) and can be self-hosted; Relace Apply 3 is proprietary.
Which is newer?
GLM-4.6 is the newest, released Sep 30, 2025. Relace Apply 3 came out Sep 26, 2025; Kimi K2 Instruct 0905 came out Sep 5, 2025. Knowledge cutoff: GLM-4.6 Apr 2025.
How do you decide the winner?
Each model gets a 0–100 score on capability (50%, independent benchmark results); price (25%, blended price per million tokens (3 input : 1 output), log scale); inputs & features (15%, image, PDF, audio and video input, tool calling, structured output and reasoning); context window (10%, maximum tokens per request, log scale). Dimensions missing for any model are dropped and the remaining weights rescaled, so every model is judged on the same evidence. Specs and prices come from public model listings and the labs’ own API pages; capability scores come from independent benchmark runs. Data updated Nov 2, 2026.
